First, the lateral PFC is a large expanse of cortex and the drugs were injected virtually in the middle of it. Second, the effects began immediately during the long, slow, injection. If the site of action were elsewhere, it would take time for the drug to diffuse to those sites. Instead, we observed that the effects began a few minutes after the start of the injection. Third, not all sites in the lateral PFC produced an effect, as might be expected if the drug had widespread actions (see Experimental
Procedures); most vlPFC, but not dlPFC, sites resulted in impairment. There are similar levels of D1R expression in these areas (de Almeida et al., 2008), so this may reflect the greater number of CB-839 price vlPFC than dlPFC neurons with selectivity during conditional visuomotor
tasks (Wilson et al., 1993), and thus suggests the vlPFC as the site of action. Finally, while SCH23390 has a preference for binding to D1Rs, it also binds LGK-974 in vivo to serotonin 5-HT2 and 5-HT1C receptors (Bischoff et al., 1988; Alburges et al., 1992). Thus, we cannot disregard the possibility that some of the SCH23390 effects are mediated in part by these receptors. The specificity of D1 receptors could be established by injecting several 5-HT antagonists, but conducting these experiments would be difficult in monkeys. Importantly, our neurophysiological results are in line with previous studies in which D1R specificity could be established. For example, a previous study in monkeys (Sawaguchi and Goldman-Rakic, 1994) showed that local injections of ketanserin, a selective antagonist of 5-HT2 receptors, near PFC sites affected by SCH23390 failed to induce any clear changes in the monkeys’ performance, suggesting that SCH23390 was acting on D1Rs. Studies in humans indicate that prefrontal D1Rs are involved in working memory (Robbins, 2000). Müller et al. (1998) demonstrated performance-enhancing effects of a mixed D1-D2 agonist on spatial working
memory, but no effect of a selective D2 agonist, suggesting a role for D1Rs. Studies in monkeys and rodents have shown that Thymidine kinase modulation by D1Rs in the dlPFC and prelimbic cortex during spatial working memory follows an inverted-U-shaped curve: too little or too much D1R stimulation causes cognitive impairment, while moderate levels of D1R stimulation strengthen and sculpt selectivity to optimize PFC function (Williams and Goldman-Rakic, 1995; Zahrt et al., 1997; Seamans et al., 1998; Granon et al., 2000; Chudasama and Robbins, 2004; Vijayraghavan et al., 2007). Our results suggest that prefrontal D1Rs in the monkey vlPFC play a relevant role in associative learning, perhaps by sculpting neural selectivity of prefrontal neurons. D1Rs in the lateral PFC may play a role in cognitive flexibility as well.